Prediction Model for Familial Aggregated HBV‐Associated Hepatocellular Carcinoma Based on Serum Biomarkers

ABSTRACT Background Accurate assessment of the risk of familial aggregated hepatitis B virus (HBV)‐associated hepatocellular carcinoma (HCC) and regular surveillance for these patients at high risk may be valuable to reduce the occurrence and improve the prognosis of HCC. Aim This study aimed to dev...

Full description

Saved in:
Bibliographic Details
Main Authors: Linmei Zhong, Guole Nie, Qiaoping Wu, Honglong Zhang, Haiping Wang, Jun Yan
Format: Article
Language:English
Published: Wiley 2025-06-01
Series:Cancer Reports
Subjects:
Online Access:https://doi.org/10.1002/cnr2.70253
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849433450110517248
author Linmei Zhong
Guole Nie
Qiaoping Wu
Honglong Zhang
Haiping Wang
Jun Yan
author_facet Linmei Zhong
Guole Nie
Qiaoping Wu
Honglong Zhang
Haiping Wang
Jun Yan
author_sort Linmei Zhong
collection DOAJ
description ABSTRACT Background Accurate assessment of the risk of familial aggregated hepatitis B virus (HBV)‐associated hepatocellular carcinoma (HCC) and regular surveillance for these patients at high risk may be valuable to reduce the occurrence and improve the prognosis of HCC. Aim This study aimed to develop a simple and reliable prediction model for the risk of HCC in these patients. Methods and Results This study analyzed clinical laboratory results from a database of 1285 patients with familial aggregated HBV who attended the First Hospital of Lanzhou University from January 2010 to December 2019. Univariate and multivariate logistic regression (LR) analysis showed that hemoglobin (Hb), neutrophil percentage (NP), total protein (TP), glutamyl transpeptidase (GGT), alglucosidase alfa (AFU), aspartate aminotransferase (AST) to Alanine aminotransferase (ALT) ratio (AAR), and alpha‐fetoprotein (AFP) were identified to be independent risk factors for HBV‐associated HCC. Prediction models were developed using a multivariate LR model, classification and regression tree, Native Bayes, Bagged tree, AdaBoost, and random forest. We used a multivariate LR model as a benchmark for performance assessment (AUC = 0.737). The results showed that the Native Bayes model had an AUC of 0.749, which was better than that of the other models. Conclusion Finally, the Native Bayes model demonstrated better predictive performance for HCC, which helped in the clinical decision‐making and identification of HCC high‐risk groups.
format Article
id doaj-art-db1ad25c4eba41fe84eff44946acbb09
institution Kabale University
issn 2573-8348
language English
publishDate 2025-06-01
publisher Wiley
record_format Article
series Cancer Reports
spelling doaj-art-db1ad25c4eba41fe84eff44946acbb092025-08-20T03:27:01ZengWileyCancer Reports2573-83482025-06-0186n/an/a10.1002/cnr2.70253Prediction Model for Familial Aggregated HBV‐Associated Hepatocellular Carcinoma Based on Serum BiomarkersLinmei Zhong0Guole Nie1Qiaoping Wu2Honglong Zhang3Haiping Wang4Jun Yan5Postgraduate Training Base Alliance, Wenzhou Medical University Wenzhou Zhejiang People's Republic of ChinaDepartment of Colorectal Hernia Surgery Binzhou Medical University Hospital Binzhou Shandong People's Republic of ChinaDepartment of Pediatrics The First Affiliated Hospital of Fujian Medical University Fuzhou Fujian People's Republic of ChinaThe First School of Clinical Medicine, Lanzhou University Lanzhou Gansu People's Republic of ChinaThe First School of Clinical Medicine, Lanzhou University Lanzhou Gansu People's Republic of ChinaThe First School of Clinical Medicine, Lanzhou University Lanzhou Gansu People's Republic of ChinaABSTRACT Background Accurate assessment of the risk of familial aggregated hepatitis B virus (HBV)‐associated hepatocellular carcinoma (HCC) and regular surveillance for these patients at high risk may be valuable to reduce the occurrence and improve the prognosis of HCC. Aim This study aimed to develop a simple and reliable prediction model for the risk of HCC in these patients. Methods and Results This study analyzed clinical laboratory results from a database of 1285 patients with familial aggregated HBV who attended the First Hospital of Lanzhou University from January 2010 to December 2019. Univariate and multivariate logistic regression (LR) analysis showed that hemoglobin (Hb), neutrophil percentage (NP), total protein (TP), glutamyl transpeptidase (GGT), alglucosidase alfa (AFU), aspartate aminotransferase (AST) to Alanine aminotransferase (ALT) ratio (AAR), and alpha‐fetoprotein (AFP) were identified to be independent risk factors for HBV‐associated HCC. Prediction models were developed using a multivariate LR model, classification and regression tree, Native Bayes, Bagged tree, AdaBoost, and random forest. We used a multivariate LR model as a benchmark for performance assessment (AUC = 0.737). The results showed that the Native Bayes model had an AUC of 0.749, which was better than that of the other models. Conclusion Finally, the Native Bayes model demonstrated better predictive performance for HCC, which helped in the clinical decision‐making and identification of HCC high‐risk groups.https://doi.org/10.1002/cnr2.70253familial aggregated HBVhepatocellular carcinomamachine learningrisk prediction
spellingShingle Linmei Zhong
Guole Nie
Qiaoping Wu
Honglong Zhang
Haiping Wang
Jun Yan
Prediction Model for Familial Aggregated HBV‐Associated Hepatocellular Carcinoma Based on Serum Biomarkers
Cancer Reports
familial aggregated HBV
hepatocellular carcinoma
machine learning
risk prediction
title Prediction Model for Familial Aggregated HBV‐Associated Hepatocellular Carcinoma Based on Serum Biomarkers
title_full Prediction Model for Familial Aggregated HBV‐Associated Hepatocellular Carcinoma Based on Serum Biomarkers
title_fullStr Prediction Model for Familial Aggregated HBV‐Associated Hepatocellular Carcinoma Based on Serum Biomarkers
title_full_unstemmed Prediction Model for Familial Aggregated HBV‐Associated Hepatocellular Carcinoma Based on Serum Biomarkers
title_short Prediction Model for Familial Aggregated HBV‐Associated Hepatocellular Carcinoma Based on Serum Biomarkers
title_sort prediction model for familial aggregated hbv associated hepatocellular carcinoma based on serum biomarkers
topic familial aggregated HBV
hepatocellular carcinoma
machine learning
risk prediction
url https://doi.org/10.1002/cnr2.70253
work_keys_str_mv AT linmeizhong predictionmodelforfamilialaggregatedhbvassociatedhepatocellularcarcinomabasedonserumbiomarkers
AT guolenie predictionmodelforfamilialaggregatedhbvassociatedhepatocellularcarcinomabasedonserumbiomarkers
AT qiaopingwu predictionmodelforfamilialaggregatedhbvassociatedhepatocellularcarcinomabasedonserumbiomarkers
AT honglongzhang predictionmodelforfamilialaggregatedhbvassociatedhepatocellularcarcinomabasedonserumbiomarkers
AT haipingwang predictionmodelforfamilialaggregatedhbvassociatedhepatocellularcarcinomabasedonserumbiomarkers
AT junyan predictionmodelforfamilialaggregatedhbvassociatedhepatocellularcarcinomabasedonserumbiomarkers